Idowu Adewumi

Mr. Idowu Adewumi

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 September 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
IAIdowu Adewumi
Idowu Adewumi
IAIdowu Olugbenga Adewumi
Idowu Olugbenga Adewumi
MSModupe Oluwayemisi Sodamola
Modupe Oluwayemisi Sodamola
Zoonotic diseases represent approximately 60-70% of new infectious diseases globally, resulting in yearly economic losses surpassing USD 120 billion from trade limitations, livestock deaths, and decreased productivity. Conventional veterinary surveillance systems, depending on manual reporting and lagging diagnostics, frequently identify outbreaks 10-14 days post-emergence, causing swift pathogen transmission. This research utilized predictive machine learning (ML) models on a synthesized dataset of 200,000 veterinary records that combined clinical, genomic, environmental, and climate factors. Supervised classifiers such as Logistic Regression, Decision Trees, Random Forests, Gradient Boosting (XGBoost, LightGBM, CatBoost), SVMs, and k-NN were assessed in both binary outbreak classification and multi-class risk prediction tasks. Random Forests attained the best AUC of 0.95, demonstrating 91% sensitivity and 88% specificity, cutting outbreak detection delay by 12 days relative to baseline reporting. Gradient Boosting models showed similar performance, achieving AUC values ranging from 0.93 to 0.94, especially standing out in structured surveillance information. Analysis of feature importance revealed that serology IgG (12.9%), antimicrobial resistance indicators (9.8%), and microbiome diversity metrics (8.5%) were the leading predictors, with climate factors accounting for another 7–10% of the variation in predicting vector-borne diseases. Case studies showed that ML models predicted avian influenza outbreaks in poultry with 92% accuracy, identified rabies in domestic and wild reservoirs with a 14% false positive rate, and forecasted brucellosis and bovine tuberculosis risks in cattle with a Cohen’s κ of 0.87, indicating strong alignment with expert assessments. The expansion of ticks and mosquitoes influenced by climate was forecasted with an average error of ±6.2% in three areas, highlighting the effectiveness of the models for proactive One Health monitoring. The results underscore how ML can improve veterinary diagnostic processes, enhance outbreak readiness, and bolster interdisciplinary One Health partnerships. In spite of issues related to data imbalance, generalizability, and restricted use in low-resource environments, predictive ML frameworks show significant promise in decreasing detection times by more than 70%, cut economic losses by USD 20-25 billion annually, and inform evidence-based veterinary policy.

24 October 2025 in Research Square
Verified
Preprint
Open access
PDF file
IAIdowu Adewumi
Idowu Adewumi
This study outlines the creation, implementation, and assessment of a Blockchain–AI integrated chain-of-custody (CoC) framework for managing digital forensic evidence. The research sought to improve the integrity of evidence, transparency, and automation, tackling the shortcomings of conventional manual CoC processes. The suggested system was executed utilizing Hyperledger Fabric (6 nodes, PBFT consensus) and Ethereum testnet (10 nodes, PoA consensus), attaining an average block time of 1.2–3.8 seconds and transaction latency of 85–150 milliseconds. Smart contracts, RegisterEvidence(), VerifyCustody(), AccessGrant(), and LogActivity() streamlined the custody procedure, achieving a 99.6% integrity validation rate in blockchain-only mode and a complete 100% validation when paired with AI anomaly detection. The AI subsystem utilized a CNN–LSTM combined model that was trained on 500 labeled transaction logs, achieving 97.2% accuracy, 0.96 precision, 0.97 recall, and an F1-score of 0.965. Correlation analysis indicated a robust positive association (r = 0.94) between AI anomaly detection and blockchain integrity verification. Scalability evaluations over 100–5,000 transactions demonstrated throughput between 135 and 80 transactions per second (TPS), while memory usage rose from 32% to 77%, verifying effective resource utilization. The system exhibited strong alignment with SDG 16 (Peace, Justice, and Strong Institutions) and SDG 9 (Industry, Innovation, and Infrastructure), achieving scores of 0.98 for transparency, 0.95 for accountability, 0.96 for innovation, and 0.94 for digital infrastructure. Comparative benchmarks indicated significant enhancements compared to baseline CoC systems: +13.1% in integrity validation, + 60.7% decrease in latency, + 97.2% increase in accuracy, and + 50% improvement in scalability. These empirical findings confirm that the Blockchain–AI framework provides a secure, transparent, and smart forensic environment, capable of revolutionizing judicial evidence handling and enhancing institutional trust via automated, data-driven processes.

6 February 2026 in Community and Social Development Journal
Verified
Journal Article
Open access
PDF file
IAIdowu Adewumi
Idowu Adewumi
Evaluation of forecasting models with 5000 daily sales records (2020-2023) from A&Z Supermarket located at Oyo State, Nigeria. Ten (10) product IDs from five (5) stores. The dataset incorporates a total of 11 predictive variables. This includes price, competitor price, macroeconomic index (simulated), promo indicators, holiday flags and rolling averages. The forecast models are assessed through a train/test split with a 70/30 ratio. They are forecasting for a 7-day horizon. Further, rolling window validation prevents leakage. According to descriptive analysis, the average sales quantity resulted in 26.49 sales units with a standard deviation of 11.58. The day of sales displays cumulative spikes, mainly during weekends as well as holidays. The research compares various traditional models such as ARIMA, SARIMA, and Linear Regression with machine learning models like Random Forest, XGBoost, LSTM, and GRU and has a RMSE ranging from 10.55 to 14.36 and R² values ranging from %u221A0.740 to 0.884. Bagging, boosting, stacking techniques are used to improve the performance of predictive models. The stacking ensemble achieved the lowest RMSE (8.14). The performance of LSTM was observed to be best as depicted by (RMSE= 8.75, R²= 0.91) as it outperformed Prophet and XGBoost. The lack of generalizability can be attributed to a narrow dataset that incorporates only a single retailer and a simulated macroeconomic index. An analysis of the business KPIs indicates a decline in inventory costs of 17.8% and of stockouts by 15.4%. Such improvements along with a 9.5% increase in ROI are encouraged through the use of LSTM (Long-Short Term Memory) investments. The results indicate that hybrid ensemble-time-series models are superior to conventional methods and offer important practical conveniences in optimizing inventory and forecasting demand in retail settings in emerging markets.

27 August 2025 in Research Square
Verified
Preprint
Open access
PDF file
IAIdowu Adewumi
Idowu Adewumi
This study presents a multi-model computational framework for predicting cholera outbreaks using spatio-temporal, climatic, and socio-environmental predictors across regions with recurrent epidemics. The dataset included 17,842 records of regions and days from January 2024 to May 2025, divided into 70% for training, 15% for validation, and 15% for testing in cross-sectional models, along with rolling-origin splits for time-series models. Two forecasting tasks were examined: (i) prediction of reported cases and (ii) categorization of outbreak severity into low (0–9 cases), medium (10–29 cases), and high (≥ 30 cases). Baseline statistical evaluations utilized Poisson and Negative Binomial regression methods. Overdispersion tests (variance/mean ratio = 2.7) highlighted the advantages of Negative Binomial models, which identified rainfall (IRR = 1.18, 95% CI: 1.10–1.26) and water salinity (IRR = 1.11, 95% CI: 1.06–1.16) as major contributors to outbreak risk, whereas sanitation coverage lowered incidence rates by 23% (IRR = 0.77, 95% CI: 0.71–0.84). Experiments with machine learning demonstrated significant enhancements in performance. Random Forest regression lowered RMSE from 41.2 (baseline) to 28.9, whereas classification reached a macro-F1 of 0.81. XGBoost enhanced classification results with macro-F1 = 0.87 and ROC-AUC = 0.91, surpassing Random Forest (macro-F1 = 0.79, ROC-AUC = 0.86). SHAP analysis identified rainfall, sanitation, and mobility index as the three primary factors, responsible for 62% of the variance in predicting outbreaks. Deep learning utilizing Long Short-Term Memory (LSTM) networks delivered the most precise time-based predictions. For a 7-day forecast, LSTM produced RMSE = 25.3 ± 6.2, MAE = 18.4 ± 4.7, and MAPE = 12.8 ± 3.1, while ARIMA showed RMSE = 27.9 ± 7.4 and MAPE = 17.5 ± 4.5, and naive benchmarks had MAPE ≥ 20%. Over a 14-day period, LSTM maintained its advantage with RMSE = 39.5 ± 10.2 and MAPE = 20.5 ± 5.6, surpassing ARIMA (RMSE = 41.2 ± 11.0; MAPE = 24.7 ± 6.3). Federated learning trials involving 5 regional clients showed performance comparable to centralized learning, achieving an accuracy of 0.84 (without differential privacy) and 0.78 (with DP, σ = 1.0). Privacy-utility trade-offs resulted in ε = 3.1–7.8 for δ = 1e-5, confirming practicality in low-bandwidth settings (average communication overhead = 11.4 MB per round). The results indicate that LSTM-based forecasting increases epidemic prediction accuracy by as much as 25% compared to ARIMA and 35% compared to naive methods, while XGBoost boosts outbreak severity classification by 8% relative to Random Forests. Federated models guarantee privacy-focused scalability with merely 5–9% loss in utility. These findings highlight the promise of combining ensemble learning, deep temporal models, and federated AI to create resilient, data-sovereign public health surveillance systems for areas susceptible to cholera.

30 September 2024 in Journal of Agribusiness and Rural Development
Verified
Journal Article
Open access
PDF file
NSNurudeen Afolabi Sofoluwe
Nurudeen Afolabi Sofoluwe
MOMaryam Ademidun Ologbonsaye
Maryam Ademidun Ologbonsaye
BOBukonla Grace Osisanwo
Bukonla Grace Osisanwo
IAIdowu Adewumi
Idowu Adewumi
This study examines the contribution of social capital to food consumption expenditure as a measure of...

27 January 2023 in Research Square
Verified
Preprint
Open access
PDF file
IAIdowu Adewumi
Idowu Adewumi
Mimosa pudicais among the medicinal herbal plant. It is a creeping annual or perennial herb often grown for its curiosity values; as the compound leaves fold inward and drop when touched and reopens within minutes later. It belongs to Fabaceae family. The general objective of these research work is to produce biofuel from sensitive plant using existing ethanol production plant. The sensitive plant was gotten from bushes at Moor Plantation, samples as used include 78% concentration of Hydrogen tetraoxosulphate (iv) acid (H2SO4), 3L, 6L and 9L of water was added with the mass of sensitive plant. The evaluation was carried out using the mass of dried sensitive plant (Mimosa pudica) such as replicate mass of sensitive plant (kg), volume of H2SO4 (L), volume of H2O (L) added to ferment the feed stock, mass of ethanol produced after been distilled (kg), volume of ethanol produced (L), time and temperature used from each ethanol extraction. Also, it was observed that the average quantity of mean of mass of ethanol produced after been distilled from 0.5kg of sensitive plant (Mimosa pudica) to be 0.0404kg, average quantity of mean of volume of ethanol produced is 0.0454 L and the average quantity of mean of temperature is 74.680C. It was concluded that the temperature of the ethanol produced increases as the mass of sensitive plant increases and its time of production of each replicate (0.5kg, 1.0kg, 1.5kg) also increases.